The Hearing Institute will soon welcome a new team led by Dr. Keith Doelling. We offer you a presentation of Keith and his project by Keith himself. Having been a familiar face at the Institute for a few years since completing a post-doctorate here, he will remain with us for the next five years.
Who are you?
Keith Doelling: I am an immigrant to France hailing originally from Berkeley, California in the United States. I spent my early years mostly playing classical and jazz music, playing tennis and video games, and studying. In my undergraduate degree at Harvard University, I studied how to reduce emotion in fearful memories, as well as the interaction between music training and dyslexia. I, then, moved to New York and studied for my PhD in Cognitive Science at New York University where I tested computational models of neural dynamics of speech and music perception. During this time, I started a funk band of neuroscientists called Debunk and started a social group for Black PhD students at NYU, spawned by the police shooting of Michael Brown in Ferguson, Missouri.
I moved to Paris to work with Dr. Luc Arnal at the Hearing Institute 6 months before the Covid Pandemic. I developed IdA’s first platform for human behavioral and electrophysiological research. Since, then, I have further developed computational approaches to support our understanding of human auditory perception, in the domains of sensorimotor interactions, music perception, and clinical audiology.
Can you present your project?
KD: My project aims to use computational models, like Deep Learning and Bayesian models, to inform our understanding of how humans integrate information in sounds.
The human auditory experience relies fundamentally on processing and integrating sequences of information. For communication for example, the ideas that we transfer from one person to another are complex enough that a single vocal utterance would be insufficient to carry the necessary information, instead we transmit sentences whose information is stored both in the list of words that are generated but most critically in their order as well.
The perception of a single item is complex enough and is processed across many features (e.g, pitch, phonemes, semantics and grammatical form). How are these multiple features tracked across the sequence and integrated to generate a whole percept? How do deficits to the sensory pathway disrupt this process?
The goal of our project will be to track the neural encoding, maintenance and prediction of these many features in the complexity of natural perception using computational approaches to analyze noninvasive human electrophysiological recordings. We will use our findings to better understand how this complex system adapts and compensates as a result of both sensory disorders (i.e. hearing loss) and neural disorders related to sequence processing such as Autism.
I am very excited to be a part of this endeavor and to search for ways to connect with other teams.
Your dreams about this project?
KD: The project has potential to revolutionize both how we study human perception in cognitive neuroscience and how we diagnose perceptual and cognitive issues in the clinic.
In both fields, success has been found by drastically reducing stimulation to short and simple parameters to isolate specific features of perception. While this approach has allowed us to make progress, it does not actually reflect human experience and therefore studies a pared down version of neural processing. This can lead to critical gaps, where patients experience symptoms in their daily lives that are not well characterized by current diagnostic tools, or critical neuronal mechanisms are missed that only appear with greater complexity.
By developing the computational approaches necessary to track sequential information in complex naturalistic experience, I aim to solve this gap not only for my own research but in a manner that can be shared with and influence the rest of the clinical and fundamental neurosciences.
Anything you would like to add?
KD: I would like to express my gratitude to all the members of our Institute for having been so welcoming to me as a foreigner and immigrant to France, during my time at the Institute. Especially, to Luc Arnal, Diane Lazard and the rest of the ACC team who have been an amazing foundation and source of support as I develop my new lines of research. As a whole, the IdA community is a big reason for why I’m excited to stay and take this new adventure. I look forward to many further opportunities to interact, to discuss science and life, and to collaborate over the next 5 years and beyond.